It’s clear that the renewables revolution will include more dispersed generation, including some by users ranging from you and me to Walmart. Besides, clever load management can quickly lop off a useful chunk of high-carbon peak generation. This dispersed, interactive grid will involve big changes in the way electricity is delivered to consumers whether we like it or not. It will be even more an exercise in social engineering than in electrical.
The documents I have are long on “strategic visions” and pretty coloured charts, short on plans and standards; there’s a long way to go. But it’s never too early for consumers to lay down what we want and what we can’t live with. That’s my standpoint here: a representative dumb consumer.
There are two really bad ideas out there.
From New Scientist:
Using information fed in by the appliances, combined with predictions of renewable power output based on local short-term weather forecasts, the [grid] operating system can balance demand to match supply by telling non-essential appliances to switch themselves off.
I’m a pretty statist European, but I’m just not prepared to accept so great a loss of our family’s control over our own house. In Colorado they’ll say “no” with lawsuits; in Idaho, I fear with guns. The scheme is Orwellian, and too vulnerable to abuse by utilities (think Enron) and malicious hackers.
The opposite idea, called dynamic demand, is to put all the intelligence in appliances. From the same article:
As the frequency drops, a built-in controller in each fridge detects the change [in voltage], checks the temperature of the fridge, and calculates how long it can stay chilled without drawing any power. It then switches the fridge off for as long as is safe.
It’s hard to know where to begin with this.
* The grid managers are trying their best to prevent voltage fluctuations. It’s asking for trouble to optimise one part of the system around failures in another part.
* We need to switch things off when carbon emissions are high. This on the face of it uncorrelated with management problems on the grid, except when it’s saturated. Compare a stormy winter night, with wind turbines going full blast, but varying with the gusts; and a still summer day when marginal electricity is supplied from big, stable gas turbines.
* I for one am not prepared to programme 20 appliances separately. My current domestic stock is listed at the end, with my preferences as to load shedding and time-shifting; yours will be different but are unlikely to be less complicated.
Therefore: Doorway negotiation
So the smartness has to consist in a negotiation between my house and the grid. The centrepoint will be a smart meter-cum-controller – an electronic janitor – connected to a home automation (domotic) network. The “janitroid” should be able to time-shift or temporarily interrupt appliances based on previous instructions from the homeowner and information from the grid. (© RBC? But the word already has 7 hits on Google so we probably can’t claim a remunerative copyright. Sigh.)
On the house side of the gateway:
* The domotics side of this is pretty well developed. There are several competing standards and technologies; the simplest looks to be one that uses existing power circuits. It will be essential to settle quickly on a single worldwide standard for the interface with the appliance controller chips. A domotic network will have other functions, like managing alarms, temperature, and home entertainment; but power management is the task that has requires wider networking and has a broad social effect.
* The programming of the disconnection or time-shifting of appliances has to be centralised in the janitroid. The US Energy Department has this one right: two hours a year, no more, is a reasonable effort to ask from Joe and Jane Average. Given the struggles most people have with programming VCRs, a lot of work has to be put into the user interface.
* The janitroid should be owned by Joe and Jane, not their utility company. It will be as important a part of our property as the locks on our doors. Alternatively janitroids could be leased from regulated independent third-party suppliers. They should be built and certified to high design standards of security and reliability: like ATMs, which are trusted by both banks and their customers.
On the side of the grid, I’ll limit myself to the question what information it should send to the janitroid. The danger is to be too clever; it may be technically possible to send a continuously variable price signal, but unlike Alcan homeowners won’t be able to respond intelligently to so much information. So keep it simple, stupid. I suggest:
* Three pricing bands: low, medium, high. For the car recharger say, the consumer could specify the low-band only; for the air conditioner, low and medium. The time bands could shift seasonally in some straightforward way.
* A continuously variable signal of marginal system carbon intensity, which green consumers could choose to display, like an MPG counter in a car, or use to switch on interruptible appliances.
* A system overload signal. This could – but only by prior contract – require the janitroid to shed a specified load. How it did this would depend on the homeowner’s preferences (a priority list).
In addition, you would need to specify the feed-in tariffs.
For this to happen in crisis rather than geological time, a lot more leadership will be needed than Bush’s Energy Department or the EU Commission have been envisaging. The crucial issue is technical standards: chip controllers. janitroids, signals from the grid. Governments shouldn’t set standards, but they can light fires under the relevant engineering committees.
Steven Chu as Energy Secretary (a great choice) could do worse than start by collecting a few thousand lists like this, to get a sense of the citizenry’s readiness to cooperate in making energy savings.
Here’s a list of the electrical appliances currently permanently plugged in in my and Pat’s house. “Interruptible” means for an hour or so, not all day. The many small appliances plugged in occasionally (coffee-maker, toaster, hair-dryer, etc) can’t generally be time-shifted or interrupted.
Chu is a Nobel prizewinning physicist. I thought I’d help him out by translating the argument for three pricing bands into simpler terms in his language:
And finally we divide by 2 to get the postulated 3.
(With apologies to Edward Witten)